Intercom, the Irish-founded tech company known for its cutting-edge customer service solutions, recently decided to make a major shift in AI choice.
They transitioned from using OpenAI’s ChatGPT to harnessing the power of Anthropic’s Claude for their AI-driven customer service tool, Fin 2.
This change was not a result of ChatGPT failing them but a deliberate step forward to enhance efficiency and effectiveness through advanced language models.
To fully understand the impact, Intercom conducted what they call a “Torture Test”. This involved rigorous evaluations to test the AI’s accuracy and problem-solving capabilities.
The result? A game-changing shift to Claude has enabled Intercom to handle twice the customer support requests without increasing their team size, achieving an impressive 99.9% accuracy in resolving 51% of queries.
This transition is part of a larger $100 million investment by Intercom to enhance their AI capabilities. They are also creating 75 new roles focused on machine learning, demonstrating their firm commitment to using AI for tackling complex customer queries and ensuring accurate responses.
So, why choose Claude from Anthropic? Anthropic emphasizes ethical AI, known as “constitutional AI”, focusing on safety and transparency.
This aligns with Intercom’s need for technology that operates on sound principles, aiming for more human-like interactions and reducing “hallucinations”, where AI may produce wildly incorrect responses.
With companies like Intercom embracing generative AI for customer service, incredible transformations are on the horizon.
Large language models are reshaping customer service into something more vibrant, efficient, and automated, benefitting both businesses and their customers.
